International Business Review 22 (2013) 450–465

Contents lists available at SciVerse ScienceDirect

International Business Review

jo u r nal h o mep age: w ww.els evier .c o m/lo c ate / ib us r ev

When does FDI matter? The roles of local institutions and ethnicorigins of FDI

Danny T. Wang a,*, Flora F. Gu b, David K. Tse c, Chi Kin (Bennett) Yim c

a Department of Marketing, Hong Kong Baptist University, Kowloon Tong, Hong Kongb Department of Management and Marketing, Hong Kong Polytechnic University, Hung Hom, Hong Kongc School of Business, The University of Hong Kong, Pokfulam Road, Hong Kong

A R T I C L E I N F O

Article history:

Received 31 May 2011

Received in revised form 27 March 2012

Accepted 13 June 2012

Keywords:

China

Emerging economies

FDI impacts

Foreign direct investment

Institutional development

A B S T R A C T

How foreign direct investment (FDI) affects a host environment is a much discussed yet

less understood topic of salience for international business managers, policy makers and

researchers. Using panel data from 287 Chinese cities over the period 1999–2005, our

study assesses (1) the multiple impacts of FDI in both positive and negative domains, (2)

the role of local institutional development in moderating these impacts, and (3) whether

the moderating effects of institutions differ depending upon the origins of the incoming

investment (ethnic- versus non-ethnic-linked). Our analysis shows that indeed, FDI is a

double-edged sword: it enhances the host city’s economic growth, labor productivity and

innovation but it also causes employment reduction and pollution in host cities. Moreover,

the host city’s institutional development is found to enhance the positive impacts of FDI

and reduce its negative ones. Interestingly, the moderating effect is smaller for ethnic-

linked FDI than for non-ethnic-linked FDI. As the first comprehensive attempt to unravel

the role of institutional development in moderating the ambiguous impacts of FDI in

multiple domains, this study confirms that a host’s ability to absorb the benefits of FDI

while curtailing its associated costs is both plausible and pivotal. As our world becomes

flatter and FDI more entrenched in a host’s economic and social development, this study

provides important implications.

� 2012 Elsevier Ltd. All rights reserved.

1. Introduction

In international business (IB), the research on foreign direct investment (FDI) has generated an extensive body ofknowledge about the critical and diverse impacts that FDI may exert on the economic and social development of hostcountries. Both positive impacts such as impacts on the host countries’ productivity and economic growth and negativeimpacts such as those on the natural environment have attracted much academic attention. However, findings in this areaare often isolated, because few studies have investigated both positive and negative impacts of FDI simultaneously. Moreimportantly, a review of existing literature often leads to inconclusive findings. For example, on productivity spillovers,while studies by Aitken and Harrison (1999) in Venezuela and Kathuria (2000) in India found a negative impact of FDI, Liu(2002) in China and Sinani and Meyer (2004) in Estonia found a positive association. On environment spillovers, the picture

* Corresponding author.

E-mail addresses: dtwang@hkbu.edu.hk (D.T. Wang), msfgu@polyu.edu.hk (F.F. Gu), davidtse@business.hku.hk (D.K. Tse), yim@business.hku.hk

(C.K. Yim).

0969-5931/$ – see front matter � 2012 Elsevier Ltd. All rights reserved.

http://dx.doi.org/10.1016/j.ibusrev.2012.06.003

D.T. Wang et al. / International Business Review 22 (2013) 450–465 451

is even less clear (Meyer, 2004). Both pollution halo (i.e. FDI reduces pollution intensity) and pollution haven (i.e. FDIaggravates pollution intensity) effects of FDI have received empirical support (Antweiler, Copeland, & Taylor, 2001; Copeland& Taylor, 2004; Dasgupta, Laplante, Wang, & Wheeler, 2002).

These inconsistencies are concerning because they suggest the lack of a theoretical framework to understand the varyingeffects of FDI (Caves, 1996; Rodrik, 1999; Wells, 1998). Consequently, our knowledge about how to hedge or reduce thenegative FDI impacts while promoting its positive ones is limited, in spite of persistent calls in the existing literature (Caves,1996; Meyer, 2004; Wells, 1998). Given the importance of the topic, there is indeed a mounting need to advance ourunderstanding about the scale and scope of the impacts of FDI on host-countries with different characteristics (Caves, 1996;Dunning & Lundan, 2008; Eden, 2009; Meyer, 2004; Narula & Dunning, 2000; Rodrik, 1999; Wells, 1998).

In an effort to address the inconsistent findings, we take a new and institution-based perspective to argue thatinstitutional environment significantly varies the degree and even the direction of FDI impacts in the host economy. Thisinstitution-based view on FDI is particularly relevant to developing countries where institutions differ significantly fromthose in developed countries and forcefully shape the way multinationals behave and interact with local sectors(Cantwell, Dunning, & Lundan, 2010; Hoskisson, Eden, Lau, & Wright, 2000; Peng, Wang, & Jiang, 2008). Our study buildson this line of research and aims to make three contributions to the FDI literature. First, our conceptualization is basedon the understanding of the institutional mechanism through which host economies absorb and exploit FDI(Borensztein, Gregorio, & Lee, 1998; Cantwell et al., 2010). We argue that local institutions in terms of private propertyprotection, legal and regulatory enforcement, product and intermediary market development, can moderate the variousimpacts of FDI on the local economy. More specifically, a more developed institutional setting motivates and facilitatesboth foreign and local firms to compete for output rationalization and curtails the negative impacts of FDI. In contrast, apoorly managed institutional setting may weaken the positive impacts of FDI while strengthening its negative ones. Byassessing the role of local institutions in moderating multiple impacts of FDI, our study enriches the institution-basedview of FDI and sheds new light on some long debated questions in IB: is FDI always good for the host economy; why dosome regions benefit from FDI more than others; and how can government policies make the impacts of FDI morefavorable (Dunning & Lundan, 2008; Meyer, 2004; Wells, 1998)?

Second, our study differentiates two types of FDI, those that are ethnic-linked and those that are not, depending on theorigin of investment (Dean, Lovely, & Wang, 2009; Wei & Liu, 2006). We examine their differential impacts across differentinstitutional settings. As ethnic-linked FDI normally shares greater similarity with domestic investment and has strongerlocal embeddedness, its effects are less subject to institutional changes. Thus, we argue that the moderating effects ofinstitutions will be less salient in promoting the positive while reducing the negative impacts of ethnic-linked FDI. Incontrast, such moderating effects are much stronger for non-ethnic-linked FDI because of its distinctly superior resourcepools and managerial practices. This fine-grained analysis of how institutions interact with FDI of different origins providesimplications to guide a more discriminating management of FDI by the host governments.

Third, our study uses Chinese cities as the empirical context because China has always been the top recipient of FDIamong developing economies since 1990. More importantly, within the country, there are around 300 major prefecture-level cities with varying institutions all competing for FDI (Qian & Weingast, 1997). Being a regionally decentralizedauthoritarian system (Xu, 2011), Chinese local governments are responsible for initiating and coordinating reforms,providing public service, enacting and enforcing laws within their jurisdictions, resulting in wide variations in institutionalquality across cities and regions. Therefore, it offers us a rich context to empirically investigate the link between institutionaldevelopment and FDI impacts in one country. We amass data from four sources over 1999–2005, representing the firstcomprehensive approach to testing the role of institutions in moderating the multiple impacts of FDI. We use two-way fixedmodels which consider both time- and city-specific effects and the instrumental variables (IV) approach that tackles theproblem of endogeneity to minimize potential bias in our results.

2. Conceptual framework and hypotheses development

Stemming from the common belief that FDI generates greater good for host economies, governments of emergingeconomies such as China give FDI a high priority on their developmental agenda. Often they offer FDI a wide array ofincentives – including investment subsidies, lower taxes, duty exemptions, and at times local market access (Meyer & Sinani,2009). In return, the FDI is expected to contribute to the local economy with positive impacts on technology transfer,management know-how, global market access, and improved industrial competitiveness (Blomstrom, 1989; Kokko, 1996).FDI is also expected to help reduce income and technology gaps between the host economy and developed countries.

A voluminous literature exists that explores the impacts of FDI on the host’s socio-economic well being (Borensztein et al.,1998; Carkovic & Levine, 2002). Yet, the findings are diverse, and sometimes far from conclusive (Meyer, 2004). In thefollowing, we discuss the theory behind the institution-based view and how the institutional settings change the scale, oreven the direction of the impacts of FDI in various socio-economic domains.

3. Institutional development of host economies

There is a growing consensus among researchers that IB studies need to recognize the role of institutions – the ‘‘rules ofthe game’’ – in shaping MNC activities and the spillover effects they produce (Cantwell et al., 2010; Dunning & Lundan, 2008;

D.T. Wang et al. / International Business Review 22 (2013) 450–465452

North, 2005). Institutions, broadly defined, consist of informal constraints such as norms, culture, and customs, or the morepurposive formal ones embodied in particular political rules and organizational structures. In this study, we focus on formalinstitutions which encompass the legal framework and its enforcement, private sector development, governmentintervention, regulatory regimes, and product and intermediary market development (Davies & Walters, 2004; North, 1990,2005; Peng et al., 2008), as these institutions are enforced constraints for FDI activities and are quantified in the existingliterature (Fan, Wang, & Zhu, 2009). In contrast, informal institutions, though playing a significant role in the revolution ofthe formal ones, emerge spontaneously and remain in the private realm (Williamson, 2000).

China provides an ideal setting to examine within-country variations in these formal institutions. Contrasting pastthree decades’ unprecedented large scale and rapid growth, China has in general been rated below average onmeasurements of the application of the rule of law or for governance quality. Such a contrast has been considered amongthe utmost challenges to the world economy by many Nobel Prize laureates, such as Coase, Maskin, Mirrlees, North, andStiglitz. Recent intellectual developments in political economy suggest that China’s distinctively different governancestructure summarized as the Regionally Decentralized Authoritarian (RDA) system (Xu, 2011) partially explains theinstitution–economics imbalance. In this regime, local governments are major players in the bulk of the Chineseeconomy. Under the supervision of the central government, they initiate, negotiate, implement, divert and sometimesresist reforms and policies. They override laws or substitute laws. It is at the sub-national level that these regionalgovernments drive, influence or hamper economic development, environmental conservation or degradation, socialstability, etc. To reflect these within-country institutional variations, we define host city institutions to be developed ifthey support the voluntary exchange underpinning an effective market mechanism. Conversely, we refer to institutionsas underdeveloped if they fail to ensure effective markets or even undermine markets.

Empirically, we make use of a NERI index compiled and updated annually by Fan, Wang, and Zhu, the National EconomicResearch Institute of China (NERI index for short) to proxy institutional development in China. The NERI index is built on 19components, covering five dimensions of institutional arrangements and policies: (1) government and market relations; (2)economic structure, in particular the growth of the non-state sector and the reform of the state enterprises; (3) inter-regionaltrade barriers, including price control; (4) factor-market development, including factor mobility; and (5) legal frameworks,including development of intermediate institutions and legal enforcement. It is the most official and comprehensivemeasure that tracks the changes of institutions in China. Higher scores of the index suggest greater institutionaldevelopment. The index has been extensively used in economics, finance, and IB studies on China (Chen, Firth, & Xu, 2009;Du, Lu, & Tao, 2008; Gao, Murray, Kotabe, & Lu, 2010; Li, Griffin, Yue, & Zhao, 2011).

Our central tenet is that the development of host institutions enhances the positive impacts of FDI on economicgrowth, labor productivity, and innovation capacity, while it weakens the negative impacts of FDI on employment and the

H4 H3

H2

H1

Overall FDI presence

Ethnic-link ed FDI (Ove rseas

investments fr om Hong Kong,

Maca u and Taiwan) in Chinese

Cities 199 9-2006

Non-ethnic -linke d FDI (Other

foreign investments) in Chinese

Cities 199 9-200 6

Positive Impact s

Economic growth

Labor productivit y

Innovation gains

Negative Impact s

Unemplo yment

Pollution I ntensity

Institutional Develo pment

Gover nment-Market Relations

Development of Non- State Econ omy

Development of Produc t Market

Development of Factors Market

Develo pment of Inte rmediary a nd Lega l Enviro nment

Control Variable s

One yea r lag GDP pe r capit a

Investment rate

Populati on growth

Labor qua lity

Gov ern ment expenditure

Labor wage

Capital inten sity Firm size R&D intensi ty Population densi ty

Industrial output

GDP per sq km

Indic ate str onger impac ts

Indic ate we aker impa cts

Fig. 1. The conceptual framework of the positive and negative impacts of FDI.

D.T. Wang et al. / International Business Review 22 (2013) 450–465 453

environment. We incorporate these five dimensions of impact into our framework because they are some of the most salientoutcomes of FDI and the primary focus of interest in the FDI literature. Moreover, our city-level data allows us to examinethese five dimensions. Note that this study is not an attempt to review exhaustively the literature on FDI impacts andreplicate the findings, which would require multiple articles. Rather, we provide a baseline model that encompasses themultiple impacts of FDI, and apply an institution-based view to reveal the underlying reason behind the variations. Due to alack of relevant data, FDI impacts on other dimensions such as corruption are beyond the scope of this study. Fig. 1 presentsour theoretical framework.

4. The role of institutions in enhancing FDI impacts

FDI contributes to a host environment’s economic growth, commonly measured by the rate of change of GDP, through anumber of avenues. In addition to the injection of new investment funds, FDI stimulates a host economy’s growth throughthe transfer of advanced technology (e.g., Sinani & Meyer, 2004) and human capital (e.g., Carkovic & Levine, 2002), creatingthe ‘‘first-order’’ effects. Subsequently, second-order effects, including the mobility of these advanced technology,management system, and skilled labor for local firms (Hale & Long, 2006) will follow. This, in turn, enhances the hostenvironment’s ability to absorb other FDI, creating clusters of FDI and pools of talented managers and a skilled labor force inthe host economy (Borensztein et al., 1998).

As for labor productivity, which refers to the value-added output per worker (Haskel, Pereira, & Slaughter, 2007), the centralpostulate is that FDI increases capital investment and improves efficiency in resource allocation, hence augmenting the laborproductivity of the host economy. FDI also offers demonstration effects, that is, their superior operational efficiency encourageslocal firms to place more emphases on technology investments for productivity gains (Meyer & Sinani, 2009). Previous workshave provided support for this positive prediction (Buckley, Clegg, & Wang, 2002, 2007; Liu, Wang, & Wei, 2009).

In comparison, the effects of FDI on innovation gains, defined as the level of patent invention in our study, are less direct,partly due to the proprietary nature and the issue of intellectual property rights in production technologies andaccompanying innovations. Yet innovation gains are known to occur through enlarged talent pools and the emergence of askilled labor force as FDI increases (Lundvall, Johnson, Andersen, & Dalum, 2002; Nelson, 1993). FDI can also benefitinnovation activity in the host country via spillover channels such as skilled labor turnovers, demonstration effects, andtechnological know-how transfer in the host market (Blomstrom, 1989; Cheung & Lin, 2004; Kokko, 1996).

However, despite the arguments and evidence in support of the positive impacts of FDI, some empirical findings suggestthe contrary, mostly because the inflows of FDI may ‘‘crowd out’’ local firms in both resource and product markets (Aitken &Harrison, 1999; Altomonte & Pennings, 2009; Spencer, 2008). By competing in the local resource market which includes land,capital and the labor force, multinational corporations (MNCs) may displace local firms, inhibiting their growth and valuecreation (De Backer & Sleuwagen, 2003). As the number of foreign entrants increases, the crowding out effect causesproduction and operating costs to rise for the local firms (Altomonte & Pennings, 2009). FDI may also crowd out local firms inthe product market as they offer superior products with stronger brand names (Kokko, 1996). Some previous research (e.g.,Kosova, 2004) argued that these crowding out effects occur in the short-run, and that FDI will produce positive impacts in thelong-run, yet this postulate of a delayed effect has not been empirically validated.

We argue that a well-developed institution that has sound regulations and incentives enhances the overall benefits of FDIon the economic growth, labor productivity, and innovation capacity of the host economy. First, where the host environmenthas more developed institutions, efficient market competition is facilitated. Local firms, unable to rely on preferentialtreatment from governments, will be more motivated to counter the challenge by raising their own efficiency (Chen, Su, &Tsai, 2007). Both local and foreign firms are encouraged to compete for output rationalization and sustainable advantageswhen institutions providing sufficient legal protection for market behavior exist (Peng et al., 2008). Under suchcircumstances, although foreign firms are likely to crowd out inefficient local firms in the initial stage of FDI inflow, localfirms will catch up soon after, leading to an overall increase in efficiency (Meyer & Sinani, 2009; Spencer, 2008).

Second, where institutions are more developed, the demonstration effects of foreign firms are stronger (Blomstrom &Kokko, 2003). The pro-business environment provides transparent regulatory systems, sufficient legal protection andservice support, assisting local managers to learn from their foreign counterparts in improving their technology, marketingand management skills (Luo & Peng, 1999). The productivity of local firms will gradually improve due to the enhancedknowledge and skills set (Cohen & Levinthal, 1990).

Third, private property is effectively protected by more developed institutions, which provides foreign and local firmswith a fundamental motivation to improve efficiency and market competitiveness (Chen et al., 2007). In aninstitutionally developed setting, firms will proactively reform organizations, re-engineer business practices, increaseemployee training, and enhance the technology content of products to maximize their own returns. The assuredentitlement to private gains pertains to innovation activities in particular. Where the market malfunctions, foreigninvestors may constrain their commitment to long-term R&D in the host setting, while local firms may take advantage ofweak intellectual property protection to attain knowledge (Khanna & Palepu, 2000). Under such circumstances, theinefficient market competition will reduce both foreign and local firms’ innovation activities. In contrast, whereinstitutions are more established, both local and foreign firms will be more engaged in developing and applyinginnovations. Foreign firms’ innovation capacity will be more easily realized in new products and processes, and throughdemonstration and competition, local firms’ innovation will also be enhanced.

D.T. Wang et al. / International Business Review 22 (2013) 450–465454

Last but not least, institutional development advances entrepreneurial dynamism which prompts FDI’s positiveexternalities – growth, productivity, and innovation gains. Recent research in business venturing in China has underscoredthe importance of changes in both formal and informal rule of games in inducing entrepreneurship (Bruton & Ahlstrom,2003; Kshetri, 2007; Peng, 2004). Formal legal institutions influencing entrepreneurship have gone through ‘‘strictprohibition, tolerance, accommodation and encouragement’’ (Peng, 2004). Society’s attitude toward business alsoincreasingly favors entrepreneurship (Djankov, Roland, Qian, & Zhuravskaya, 2006), especially related to high technologies(Harwit, 2002). However, it would be erroneous to assume the existence of a standard level of institutional development inChina. In places where institutions are more developed and market-oriented, such entrepreneur-friendly mechanisms arestronger, facilitating small and medium enterprises to compete with and learn from foreign firms. In contrast, in placeswhere private entrepreneurs lack legal protection, credit support, or social environment, it is less likely for them to grow anddevelop in the face of fierce competition that FDI brings. Overall, we posit that:

H1. Local institutional development enhances the positive impacts of FDI on (a) economic growth, (b) labor productivity, and(c) innovation gains.

Existing literature on negative impacts of FDI is comparatively thinner than its economic impacts especially for emergingeconomies (Meyer, 2004). Two streams of literature on the influence of FDI on employment (Jenkins, 2006; Rama, 2003;UNCTAD, 1994) and the natural environment (Antweiler et al., 2001; Dean et al., 2009; He, 2006) have emerged.

Along the employment dimension, FDI can increase local employment directly by creating new ‘‘greenfield’’ plants(Rama, 2003), and indirectly through spurring backward or forward linkages (Ernst, 2005; Liu et al., 2009). These‘‘crowding in’’ effects may endure if the FDI makes long-term commitments to the host economy. However, evidenceabound that FDI also creates ‘‘crowding out’’ effects on employment, which may even outweigh its positive impacts(Jenkins, 2006; Rama, 2003; UNCTAD, 1994). As discussed earlier, FDI may displace local firms, resulting in net job lossesfor the host economy. Instead of opening new operations and creating new jobs, FDI may acquire local firms, reduce theuse of manual labor and switch to automation to meet its superior efficiency. Furthermore, as FDI relocates local supplychains to overseas, employment is further reduced.

The negative net impact of FDI on employment is particularly likely in China. First, as foreign firms enter China, they putcompetitive pressure on private, collective and state-run local companies. In return, local firms follow their foreigncompetitors in restructuring and streamlining their staff so as to improve their productivity (Liu, Parker, Vaidya, & Wei,2001). This demonstration effect combined with the crowding out of inefficient firms has caused total employment to shrink(Wang & Zhang, 2005). Moreover, in terms of industry composition, capital-intensive industries such as manufacturing andservices are the favored sectors of FDI. In these industries, the direct positive increase in employment is relatively limited dueto a heavy emphasis on automation. According to China Statistical Yearbooks, FDI grew from USD 4.0 billion to 82.8 billionfrom 1991 to 2002, a growth of 2100%, yet employment created by foreign subsidiaries only rose from 1.65 million to 7.58million, representing a disproportionally lower increase of 460%.

There are two major theoretical postulates regarding the relationship between FDI and the natural environment. The‘‘pollution halo’’ hypothesis suggests that FDI transfers modern, environmentally friendly technologies and productionprocesses which improve the poorer environmental standards prevalent in the host economy, leading to reduced pollution inthe host country (Antweiler et al., 2001). On the other hand, the ‘‘pollution haven’’ hypothesis suggests that due to the profit-driven nature of FDI, foreign investors are motivated to transfer outdated technologies to places where environmentalregulations are less stringent, causing greater pollution to the environment (Leonard, 1988; Low & Yeats, 1992). Empiricalevidences on the environmental impact of FDI are mostly anecdotal or cross-sectional, partly due to the lack of pollution datain many countries (Copeland & Taylor, 2004).

In China, we believe that both pollution halo and pollution haven effects exist. As FDI brings advanced technologies andproduction processes into the country, the overall environmental standards in China should be enhanced. However, thepollution haven effect is equally, if not more, salient, for three reasons. First, as previously discussed, FDI mainly focuses onmanufacturing industries, such as transportation machinery, electrical and electronics equipment, textile, and chemicalproducts, which generate much more pollution than the agriculture and services sectors (He, 2006). Second, moreimportantly, China’s legal institutions for environmental protection are far from rigorous and consistent, which allows thepollution haven effect to prevail. Christmann and Taylor (2001) pointed out that China’s environmental regulations areflexible with wide regional variations. The interpretation and enforcement of the regulations are largely dependent on localgovernments. For example, when firms emit pollutants beyond official allowable levels, they can get away with it if theyhave a good relationship with local officials, or they can choose to pay a fine, which is considered more economical by mostfirms than to invest in ways to structurally reduce pollution. In addition to slack enforcement, variations in environmentalstandards in different regions also encourage local firms and FDI to move their operations to places where regulations areless stringent. Finally, although FDI may improve the environment through demonstration effects and technology transfers,Dean et al. (2009) have suggested that local firms may very well have little interest in doing the same.

We argue that a more refined understanding about the negative impacts of FDI may be obtained by applying theinstitution-based view. As discussed previously, the negative impact on employment is largely a result of the crowding outeffect where foreign firms displace local ones overriding the crowding in effect where employment opportunities are createdby foreign investors. In a more institutionally developed setting where local firms learn and compete more effectively, webelieve a balance between the two will emerge, or in a more optimistic situation, the crowding in effect will start to

D.T. Wang et al. / International Business Review 22 (2013) 450–465 455

dominate. The newly created job opportunities by foreign firms will compensate for the loss of jobs resulting from displacedlocal firms. Furthermore, the market-supporting institutions will facilitate the development of backward and forwardlinkages for foreign investors (Ernst, 2005), which will spur the clustering of suppliers and customers in the host city,creating more job opportunities.

In addition, the institutional development will also make FDI less threatening to the environment. Where institutions aremore developed, the regulatory system of environmental protection is more transparent, consistent, and stringent. Undersuch circumstances, the pollution halo effect will be more pronounced in that foreign investors will be more motivated totransfer environmentally friendly technologies, raising the environmental standards in the society. Moreover, given a stronginstitutional setting, the high cost attached to transgressive behavior deters foreign firms from so doing, alleviating thepollution haven effect. Finally, with the development of the host institutions, the services sector will attract more foreigninvestment compared with the more energy-consuming manufacturing sector. Therefore, we posit that:

H2. Institutional development dampens the negative impacts of FDI on (a) employment and (b) the natural environment ofthe host setting.

Previous studies have discovered that FDI with different ownership origins tend to exhibit contrasting profiles in terms ofspillover effects (Buckley et al., 2007; Dean et al., 2009; Wei & Liu, 2006). In the Chinese context, two types of FDI have beencharacterized: those that are ethnic-linked (from Hong Kong, Macau and Taiwan) and those that are not (mostly from NorthAmerica, EU and Japan). While ethnic-linked FDI possesses ownership advantages in the application of standardized andmature technologies, non-ethnic-linked FDI employs state-of-the-art technologies with heavy reliance on R&D to produceinnovative and differentiated products (Wei & Liu, 2006). Comparatively, non-ethnic-linked FDI can compete at lowermarginal cost through accessing their parents’ ownership advantages such as advanced technologies, managerial skills, andmarketing know-how. As a result, they are at a more advantageous position to capitalize on the knowledge gap with localfirms (Buckley et al., 2007). In contrast, ethnic-linked FDI may have less experience in mass production, and do not possessthe same level of organizational skills as their non-ethnic-linked counterparts. Their demonstration effect to the localeconomy is comparatively limited because their operations tend to be less technologically advanced (Davies, 1996).Employees from ethnic-linked FDI taking jobs in domestic firms also bring less advanced technologies and managementpractices than those from non-ethnic-linked FDI.

The distinct nature of these two types of FDI leads us to propose the differentiated moderating effect of institutions.As mentioned previously, ethnic-linked FDI is more similar to local firms in terms of culture, technology, andorganizational skills than non-ethnic-linked FDI. In its operations, ethnic-linked FDI also has stronger socialembeddedness in the host economy (Gu, Hung, & Tse, 2008). They emphasize on building stronger social ties withgovernment bureaucracies and expanding relational networks (guanxi) in the local context. In a sense, ethnic-linked FDIis more adaptive to different local institutions and their performance is less contingent on changes in the institutionalenvironment (Luo, 2003). Therefore, we argue that host institutions that facilitate the positive and mitigate the negativeimpacts of FDI are less likely to affect ethnic-linked FDI.

In contrast, FDI from non-ethnic-linked economies possesses superior resources and practices and has less socialembeddedness in the host economy (Dean et al., 2009; Wei & Liu, 2006). They have a greater need for market-orientedinstitutions to protect their intellectual property rights and facilitate their operation. Their impacts on the local economy arealso likely to vary in a greater magnitude due to the differing levels of absorptive capability in the host cities. In addition, thelocal institutions’ role in regulating the negative impacts of FDI is likely to be more effective for non-ethnic-linked FDIbecause they are less likely to bypass formal regulations through navigating informal networks of guanxi (Gu et al., 2008).Taken together, we posit that:

H3. The moderating effect of institutional development is weaker for ethnic-linked FDI than for non-ethnic-linked FDI inenhancing the positive impacts of the FDI on (a) economic growth, (b) labor productivity, and (c) innovation gains of thehost setting.

H4. The moderating effect of institutional development is weaker for ethnic-linked FDI than for non-ethnic-linked FDI inreducing the negative impacts of the FDI on (a) employment and (b) the natural environment of the host setting.

5. Data and methodology

5.1. Database

We test the hypotheses using a panel database compiled from four major sources on 287 Chinese cities over theperiod of 1999–2005. First, a team including the first author and three research assistants extracts the bulk of city-leveldata from the Urban Statistical Yearbooks of China, published by the National Bureau of Statistics of China. To minimizemissing values, the research team complements this data with statistics from the Provincial Statistical Yearbooks and theChina Statistical Yearbooks. Second, the team collects the mass sulfur dioxide emission data of each city from the reportsby China’s Environmental Protection Agency. Third, the team obtains each city’s approved patent application data fromthe Intellectual Property Bureau of China. Finally, the team employs the NERI index by Fan, Wang, and Zhu, and matches

D.T. Wang et al. / International Business Review 22 (2013) 450–465456

it with our panel data. This multi-source database took one year to compile so that the various impacts of FDI and themoderating role of local institutions can be systematically analyzed.

5.2. Methodology

We follow the standard procedure of previous studies (Buckley et al., 2002; Liu et al., 2001) to construct the econometricmodel:

Yitk ¼ b1FPit p þ b2MKIit þ b3FPit p � ISTit þ X0itkg þ atk þ hik þ eitk

where Yitk is the specific outcome of interest for city i in year t, with k representing the five dimensions under consideration inthis paper (k = 1, 2, 3, 4, 5). FPitp is the proportion of foreign output to total industrial output in city i for year t, with p denotingthe total share of FDI, ethnic-FDI and non-ethnic-FDI (p = 1, 2, 3 respectively). ISTit is the institutional variable measured bythe NERI index. All variables in the sample are in real terms and have undergone log transformation to ensure that thecoefficients are within the easy-to-interpret range, and to partly solve outlier and heteroskedasticity problems. The keyvariable of interest is FPitp� ISTit, which is the interaction between FDI presence and the institutional development in thecity. X0 is a vector of control variables. The control variables of each equation depend on different outcome variables. Therationale of including these control variables and their measurements will be discussed later. atk and hik are sets of time andcity dummies that control for the fixed effects of time variations and city variations in the estimation. eitk is a random errorterm reflecting omitted variables.

While it is possible to use the ordinary least squares (OLS) model on pooled cross-sectional data, it is not the optimalmethod because our sample contains repeated city-level observations over time, which give rise to within-cityautocorrelation and heteroskedasticity. To avoid the potential problems of OLS, we use the generalized least squares (GLS)regression instead. To choose between fixed-effects and random-effects models, we run Hausman tests for all the models (forresults, see Table 3). The results reject the random-effects models’ assumption that the estimated panel error is notcorrelated with the independent variables. Therefore, we use two-way fixed effects models (Wooldridge, 2002) to control foryear- and city-specific effects, which is consistent with previous research on similar topics (e.g. Aitken & Harrison, 1999;Antweiler et al., 2001; Liu et al., 2009).

Furthermore, we consider and address a few econometric problems in our data. First, FDI may be attracted to places withhigher economic growth, productivity, innovativeness, and lower employment and environmental standards. Similarly,better developed host economies are likely to have better institutions. These are potential problems that could lead toreverse causality. Second, some determinants of the outcomes of interest that are naturally correlated with FDI andinstitutions may be omitted. For example, local business culture may affect both the inflow of FDI, institutions, and the socio-economic performance of the city. That is, it is likely that covðFPit p � FPi p; eitkÞ 6¼ 0 and covðMKIit � MKIi; eitkÞ 6¼ 0 in fixed-effects models. Third, the measures of FDI presence and institutions may contain errors arising from data reporting andrecording processes. For example, as the measures of institutions are constructed ex post, the analysts may have a naturalbias in seeing better institutions in richer places. These problems could be solved with instruments for FDI and institutions(Acemoglu, Johnson, & Robinson, 2001). A good instrument would be a variable which is highly correlated with FDI andinstitutions but not with the error term in these regressions. While employing ideal IVs for our study is difficult, given thetopic on FDI and institutions,2 we have tried to alleviate the problem by using as instruments the one-year lagged values ofFDI and one-year lagged values of IST throughout our regressions, following previous studies (Alfaro, Chanda, Kalemli-Ozcan,& Sayek, 2004; Borensztein et al., 1998). The results of the IV estimation yield consistent results with the two-way fixedeffects regressions. The robust findings thus go a long way in allaying our concerns on endogeneity and reverse causality.

Finally, we check for potential serial correlation in the residuals using the test developed by Wooldridge (2002: 282–283)for panel data models. To correct for heteroskedasticity, we report the results with robust standard errors to produce validstandard errors (White, 1980).

In the following, we describe the measures of FDI, institutional development, and the control variables. Table 1 provides asummary of the measures for all variables in this study.

5.3. Measures of FDI and institutional development

Overall foreign presence (FPtotal) is measured by the proportion of overall foreign share to each city’s industrial output.Ethnic-linked FDI (FPel) is measured by the output share of investments from Hong Kong, Macau, and Taiwan, and

2 It is customary in the finance and economics literature to use such IVs as legal origin or colonial history. These variables, however, are not useful in our

case, because both our dependent variables and FDI are driven by the institutional environment. We look for variables that explain FDI but are less likely to

be related to our dependent variables such as economic growth, labor productivity, innovation gains, employment, and natural environment. The literature

provides a menu of FDI determinants, but most variables are institutional. Following the evidence provided by previous studies (e.g. Alfaro et al., 2004;

Borensztein et al., 1998; Wheeler & Mody, 1992) that FDI is self-reinforcing, i.e. existing stock of foreign investment is a significant determinant of current

investment decisions, lagged FDI is used as instrument for FDI (model 1). Likewise, we add lagged score in NERI index as instrument for institutions in

Models 2 and 3. The results in Table 3 support our main hypotheses that local institutions strengthen the positive impacts of FDI and reduce its negative

ones.

Table 1

Variables and measures.

Variable Measure Abbr.

Explanatory variables Overall FDI presence Overall foreign output share in each city (%) FPtotal

Ethnic-linked FDI The output share accounted for by investments from Hong Kong,

Macau, and Taiwan in each city (%)

FPel

Non-ethnic-linked FDI The output share accounted for by investments from other foreign

areas in each city (%)

FPnel

Institutional Development The 5-dimension index published and provided by the National

Economic Research Institute of China.

IST

A: Economic growthDependent variable GDP growth rate The rate of change in GDP per capita GDPG

Control variables Initial GDP per capita One year lag GDP per capita L.GDPpc

Investment rate The ratio of fixed investment to GDP (%) INVR

Population growth Population natural growth rate (%) POPG

Labor quality The share of college graduates in total employment LQUA

Government expenditure The ratio of budgetary expenditure of local government to GDP (%) GOVEXP

B: Labor productivityDependent variable Labor productivity The value-added industrial output per worker (yuan/person) LAPRO

Control variables Capital intensity The ratio of net value of fixed assets per worker (100 million yuan) CAPINT

Labor quality The share of college graduates in total employment (%) LQUA

Firm size The value of fixed assets per firm (yuan) FIRSI

R&D intensity The expenditure for science administration per employee (yuan) RDINT

C: Innovation gainsDependent variable Number of patent The number of approved patent applications obtained from

China’s Intellectual Property Bureau

PATENT

Control variables R&D Expense The ratio of government expenditures on science and technology

development to GDP (%)

RDEXP

Labor quality The share of college graduates in total employment (%) LQUA

Industrial output Gross industrial output value of enterprises above designated size

(100 million yuan)

INDOUT

D: UnemploymentDependent variable Unemployment rate The rate of the unemployed in the labor force UNEMPL

Control variables Firm size The value of fixed assets per firm (yuan) FIRSI

Investment rate The ratio of fixed investment to GDP (%) INVR

Labor wage Average wage of staff and workers in each city (yuan) WAGE

Industrial output Gross industrial output value of enterprises above designated size

(100 million yuan)

INDOUT

E: EnvironmentDependent variable Concentration of SO2 emission Total factory sulfur dioxide emission divided by the land area of

individual cities (tons/square kilometer)

SO2INT

Control variables Capital intensity The ratio of net value of fixed assets per worker (100 million yuan) CAPINT

Population density The number of people per square kilometer of city land area POPDEN

Industrial output Gross industrial output value of enterprises above designated size

(100 million yuan)

INDOUT

GDP per square kilometer Gross domestic product per square kilometer of city land (yuan

/square metre)

GDPSK

D.T. Wang et al. / International Business Review 22 (2013) 450–465 457

non-ethnic-linked FDI (FPnel) is measured by the output share of foreign investments other than those from Hong Kong,Macau, and Taiwan. These output-based measures help us to avoid the potential pitfalls of using the input of FDI to predictoutput variables which may generate biased results (Buckley et al., 2007).

Institutional development is measured by the NERI index over 1999–2005, which reflects the sub-national variations ininstitutional development in China. As described previously, the index consists of 19 components of institutionalarrangements and policies, categorized into five dimensions. Sub-indices of the each component were computed using datafrom the statistical yearbooks, reports from the administration of industry and commerce, survey data, etc. A score for eachprovince was given based on objective measures, such as the ratio of lawyers or the ratio of accountants to the provincialpopulation, and then normalized to a value between 0 and 10 proportionately to measure institutional conditions relative toother provinces. The final index score for each province is the weighted average of the 19 sub-indices, with higher scoresindicating a more developed institutional setting.3 As the institutional development in China is unbalanced across regionsand disparate in multiple dimensions, the NERI index is widely used to capture provincial differences in institutionaldevelopment over time (Chen et al., 2009; Du et al., 2008; Gao et al., 2010; Li et al., 2011).

3 For detailed calculation of the NERI index, please refer to Fan et al. (2009).

D.T. Wang et al. / International Business Review 22 (2013) 450–465458

5.4. Measures of dependent and control variables

5.4.1. Economic growth

To examine the impact of FDI on economic growth, we follow the existing literature to compute GDP growth rate (GDPG)as the dependent variable and include initial GDP per capita (L.GDPpc) in our analyses, because it is generally believed thatpoorer places possess greater potential for economic growth (Stock & Watson, 2003). In addition, we control for the effects ofinvestment rate (INVR) measured by the ratio of fixed investment to GDP, population growth (POPG) which may hampereconomic growth given limited resources (Li & Zhang, 2007), labor quality (LQUA) measured by the share of collegegraduates in the population which is expected to boost economic growth in the city (Borensztein et al., 1998), andgovernment expenditure (GOVEXP) measured by the ratio of budgetary expenditure of local government to GDP, which maynegatively affect economic growth (Barro, 1991).

5.4.2. Labor productivity

The value-added industrial output per worker is computed to measure labor productivity (LAPRO). This is in line withLiu et al. (2001) and Buckley et al. (2002, 2007). Our control variables for the equation include capital intensity (CAPINT)in terms of fixed assets per worker, labor quality (LQUA), firm size (FIRSI) in terms of fixed assets per firm, and R&Dintensity (RDINT) measured by R&D expenditure per employee.

5.4.3. Innovation gains

We use the number of approved patent applications (PATENT)4 to represent innovation gains in each city. This measureoffers some advantages. Previous studies that examined the impact of FDI on innovation tended to use R&D expenditure orreturn on R&D expenditure as a proxy of innovation outcome (Feinberg & Majumdar, 2001). We consider the number ofpatents to be a more direct outcome of innovation activities in a society. Also, it includes both product and processinnovation (Cheung & Lin, 2004), which cannot be captured by expenditure or sales measures. To separate the effect of FDIfrom the effects of other variables, we include three control variables following Cheung and Lin (2004): labor quality (LQUA),industrial output (INDOUT), and the ratio of government-led R&D expenses to GDP (RDEXP).

5.4.4. Unemployment rate

We define unemployment (UNEMPL) as the percentage of people in the labor force who are unemployed. The controlvariables that may affect unemployment include firm size (FIRSI), investment rate (INVR), both of which are positivelyassociated with unemployment because capital and labor are considered substitutes in production, labor wage (WAGE)which may also increase unemployment due to the rising cost of labor, and industrial output (INDOUT) which is likely todecrease unemployment because more labor is needed when the scale of the economy increases (Jenkins, 2006).

5.4.5. Pollution intensity

To measure pollution intensity of a city (SO2INT), we use data of total factory SO2 emission divided by land area (tons/square kilometer), based on the reports by China’s Environmental Protection Agency. Following previous literature (e.g.,Antweiler et al., 2001), we control for the effects of capital intensity (CAPINT), population density (POPDEN) which isnormally associated with greater pollution intensity, industrial output (INDOUT) which increases SO2 emission due to thescale effect, and GDP per square kilometer (GDPSK) in the analyses.

We report summary statistics including means, standard deviations and correlations of all variables included in the fiveequations of FDI impacts in Table 2. We further check the variance inflation factors (VIF) for all variables. For the fiveequations, the mean VIF ranges from 1.43 to 2.87, indicating that multicollinearity is not a serious problem.

6. Empirical results

6.1. Main effects of FDI

We report the results of IV regressions in Table 3. We first run separate analyses on economic growth (GDPG), laborproductivity (LAPRO), innovation gains (PATENT), unemployment rate (UNEMPL), and pollution intensity (SO2INT) withrelevant control variables and FPtotal to assess the main effects of FDI. The results are displayed under Model 1 in panels A–E,providing a baseline for comparison with subsequent moderated regressions.

The coefficients of FPtotal show statistical significance in panels B and D, but not in panels A, C, and E, suggesting that thepresence of FDI has positive influence on the host city’s labor productivity (b = .131, p < .05), but at the same time, causesunemployment (b = �.238, p < .05). Although the presence of FDI does not show direct impact on GDPG, PATENT, or SO2INT,its effect transpires through the moderating role of institutions as shown in our later analyses. Taken together, the resultsreveal that FDI generates both positive and negative impacts, confirming its multifaceted and double-edged effects.

4 We did consider the Poisson regression for the analysis of PATENT. However, as the Poisson model is mainly used for dependent variables with positive

yet low values, it is not suitable for our data, which contains the number of annually approved patents ranging from 6424 to 16,786 for large cities such as

Beijing, and from 897 to 2603 for medium cities such as Jinan.

Table 2

Means, standard deviations and correlations.

Mean s.d. (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14) (15) (16) (17) (18) (19) (20) (21) (22)

Dependent Var.(1) GDPG 0.12 0.12 1.00

(2) LAPRO 2.06 0.91 0.15 1.00

(3) PATENT 4.39 1.63 �0.01 0.53 1.00

(4) UNEMPL �2.97 0.79 0.02 0.09 �0.05 1.00

(5) SO2INT 2.89 1.74 0.11 0.46 0.37 0.02 1.00

Explanatory Var.(6) FPtotal �2.41 1.34 �0.03 0.24 0.39 �0.01 0.11 1.00

(7) FPel �3.43 1.48 �0.06 0.14 0.27 �0.08 0.05 0.74 1.00

(8) FPnel �3.06 1.45 �0.02 0.20 0.37 0.04 0.06 0.83 0.35 1.00

(9) IST 1.71 0.29 0.06 0.58 0.41 0.04 0.21 0.43 0.41 0.30 1.00

Control Var.(10) L.GDPpc 0.24 0.69 �0.01 0.74 0.69 �0.08 0.42 0.27 0.19 0.25 0.43 1.00

(11) INVR �0.99 0.46 0.18 0.02 �0.05 0.04 0.11 �0.01 0.01 �0.02 0.09 �0.03 1.00

(12) POPG 1.52 0.72 �0.04 �0.17 �0.25 �0.19 �0.05 �0.11 �0.03 �0.15 �0.14 �0.11 �0.05 1.00

(13) LQUA �4.22 1.07 0.08 0.26 0.44 0.04 0.39 0.23 0.12 0.24 0.12 0.46 0.22 �0.16 1.00

(14) GOVEXP �2.37 0.42 0.03 �0.27 �0.10 0.13 �0.04 0.03 0.03 0.02 0.01 �0.29 0.28 �0.09 �0.02 1.00

(15) CAPINT 1.54 0.73 0.12 0.77 0.32 0.11 0.40 0.08 0.03 0.05 0.30 0.58 0.07 �0.15 0.17 �0.18 1.00

(16) FIRSI �0.30 0.74 0.19 0.66 0.30 �0.03 0.44 �0.14 �0.22 �0.14 0.15 0.56 0.01 �0.04 0.27 �0.24 0.59 1.00

(17) RDINT �7.37 0.97 0.07 0.40 0.47 �0.15 0.27 0.23 0.20 0.20 0.34 0.58 0.08 �0.10 0.27 0.11 0.30 0.21 1.00

(18) RDEXP �4.30 0.51 �0.02 �0.42 �0.30 0.04 �0.28 0.04 0.07 0.03 �0.05 �0.57 0.22 �0.01 �0.23 0.64 �0.32 �0.50 �0.06 1.00

(19) WAGE 9.34 0.37 0.20 0.65 0.45 0.10 0.29 0.25 0.19 0.21 0.48 0.65 0.30 �0.16 0.41 �0.02 0.52 0.43 0.36 �0.10 1.00

(20) GDPSK 6.88 1.38 0.02 0.59 0.70 �0.03 0.66 0.31 0.25 0.23 0.44 0.73 �0.01 �0.09 0.49 �0.16 0.39 0.43 0.46 �0.46 0.39 1.00

(21) INDOUT 4.65 1.53 0.07 0.81 0.85 �0.04 0.49 0.30 0.16 0.30 0.46 0.81 �0.06 �0.22 0.41 �0.23 0.61 0.63 0.49 �0.47 0.57 0.73 1.00

(22) POPDEN 6.58 0.99 �0.09 0.27 0.49 �0.00 0.59 0.23 0.21 0.15 0.31 0.30 �0.04 �0.04 0.34 �0.03 0.11 0.17 0.21 �0.24 0.07 0.87 0.44 1.00

Note: N = 2009. Correlations greater than or equal to 0.05 are significant at p< 0.05. All variables are log-transformed.

D.T

. W

an

g et

al.

/ In

terna

tion

al

Bu

siness

Rev

iew 2

2 (2

01

3)

45

0–

46

5

45

9

Table 3

The role of institutional development in moderating FDI impacts.

Model 1 Model 2 Model 3 Hypothesis result

A: Dependent variable GDPGL.GDPpc �0.232*** (0.050) �0.531*** (0.068) �0.590*** (0.074)

INVR 0.055*** (0.019) 0.001 (0.017) �0.001 (0.019)

POPG �0.000 (0.012) �0.000 (0.012) 0.005 (0.012)

LQUA 0.128*** (0.021) 0.100*** (0.019) 0.115*** (0.022)

GOVEXP �0.148*** (0.030) �0.139*** (0.030) �0.148*** (0.032)

FPtotal 0.017 (0.020) �0.013 (0.022)

IST 0.777*** (0.099) 0.856*** (0.112)

IST � FPtotal 0.045* (0.024) H1a supported

FPel �0.002 (0.016)

FPnel �0.039 (0.030)

IST � FPel 0.001 (0.021)

IST � FPnel 0.052** (0.022) H3a supported

Model F 17.62*** 21.96*** 17.95***

Hausman test/x2 141.84*** 284.79*** 281.66***

B: Dependent variable LABPROCAPINT 0.379*** (0.050) 0.372*** (0.057) 0.404*** (0.053)

LQUA 0.050 (0.036) �0.097*** (0.028) �0.099*** (0.030)

FIRSI 0.781*** (0.047) 0.479*** (0.061) 0.473*** (0.051)

RDINT 0.031 (0.021) 0.024 (0.017) 0.028 (0.018)

FPtotal 0.131** (0.054) 0.043 (0.045)

IST 1.610*** (0.123) 1.645*** (0.119)

IST � FPtotal 0.152*** (0.046) H1b supported

FPel 0.063** (0.031)

FPnel 0.025 (0.050)

IST � FPel 0.103** (0.042)

IST � FPnel 0.109*** (0.036) H3b not supported

Model F 317.99*** 341.04*** 265.94***

Hausman test/x2 163.37*** 19.26** 22.31**

C: Dependent variable PATENTRDEXP 0.033 (0.099) 0.039 (0.097) 0.091 (0.105)

LQUA 0.171*** (0.061) �0.068 (0.064) �0.033 (0.068)

INDOUT 0.762*** (0.077) 0.287*** (0.077) 0.355*** (0.081)

FPtotal 0.197 (0.135) 0.097 (0.146)

IST 2.653*** (0.357) 2.401*** (0.371)

IST � FPtotal 0.487*** (0.128) H1c supported

FPel �0.058 (0.101)

FPnel �0.055 (0.167)

IST � FPel 0.111 (0.140)

IST � FPnel 0.352*** (0.104) H3c supported

Model F 83.97*** 80.97*** 57.54***

Hausman test/x2 27.46*** 84.43*** 75.61***

D: Dependent variable UNEMPLFIRSI 0.147 (0.101) 0.279** (0.126) 0.256* (0.131)

INVR 0.057 (0.082) 0.178* (0.096) 0.252*** (0.097)

WAGE 0.091 (0.115) 0.672** (0.323) 0.514* (0.305)

INDOUT �0.122** (0.061) 0.124 (0.125) 0.117 (0.122)

FPtotal 0.238** (0.120) 0.364*** (0.130)

IST �1.882** (0.797) �1.683** (0.820)

IST � FPtotal �0.084 (0.134) H2a not supported

FPel 0.145 (0.095)

FPnel 0.223 (0.148)

IST � FPel �0.088 (0.106)

IST � FPnel �0.149* (0.090) H4a supported

Model F 6.12*** 3.74*** 3.03***

Hausman test/x2 83.51*** 65.94*** 58.47***

E: Dependent variable SO2INTCAPINT �0.074 (0.092) �0.098 (0.094) �0.129 (0.125)

POPDEN 0.470 (0.312) 0.795** (0.395) 0.938** (0.425)

INDOUT 0.150 (0.130) 0.080 (0.150) �0.009 (0.172)

GDPSK 0.740*** (0.209) 0.411 (0.302) 0.330 (0.312)

FPtotal �0.133 (0.192) �0.248 (0.222)

IST 1.297 (0.813) 1.711** (0.743)

IST � FPtotal �0.420** (0.193) H2b supported

FPel �0.090 (0.127)

FPnel �0.299 (0.252)

IST � FPel �0.074 (0.152)

IST � FPnel �0.271* (0.162) H4b supported

D.T. Wang et al. / International Business Review 22 (2013) 450–465460

Table 3 (Continued )

Model 1 Model 2 Model 3 Hypothesis result

Model F 46.34*** 35.19*** 23.70***

Hausman test/x2 23.57*** 49.20*** 60.07***

Note.

1. The total number of observations is 2009, and the total number of groups (cities) is 287. Unstandardized coefficients are reported.

2. Robust standard errors that correct for heteroskedasticity are in parentheses.

3. The Anderson–Hsiao model with IVs is used throughout all models to control for endogeneity. In Model 1, FDI is instrumented by one-year lagged value o

FDI. Models 2 and 3 control for the endogeneity problem in both FDI and local institutions by instrumenting FDI with one-year lagged FDI and loca

institutions with the one-year lagged score of NERI index.

4. Year and city fixed effects are included but not reported.

* Significance at the 10% level.

** Significance at the 5% level.

*** Significance at the 1% level.

5 In Models 2 and 3 of panel B, labor quality shows negative impact on labor productivity (p < .01), suggesting a labor-intensive nature of the Chinese

economy.6 In Models 2 and 3 of panel D, control variables including firm size, investment rate, and wage are all positively related to unemployment rate, consisten

with our prior expectation.7 In Models 2 and 3 of panel E, population density shows a positive effect on SO2 intensity, consistent with previous findings and our expectation.

D.T. Wang et al. / International Business Review 22 (2013) 450–465 461

f

l

The coefficients of the control variables are mostly significant with expected signs. In panel A, one-year lagged GDPpcnegatively affects GDPG, confirming the conventional wisdom that poorer places pick up on economic growth morequickly (Borensztein et al., 1998). Both investment rate and labor quality have positive impacts on GDPG as expected.Government expenditure as a proportion of GDP shows negative influence, which is also consistent with findings inclassic papers (Barro, 1991). In panel B, in addition to the significant impact of FDI, control variables including capitalintensity and firm size show positive influence on labor productivity, consistent with previous findings5 (Buckley et al.,2002; Liu et al., 2001). In panel C, labor quality and industrial output are positively related to the number of patentsapproved each year. Interestingly, R&D expense by the government (RDEXP) has no significant impact. Combined withthe negative effect of government expenditure in panel A, the result suggests the relative inefficiency of the governmentin boosting economic growth or innovation.

The control variables in panels D and E also show a consistent pattern of results with previous literature. In panel D,industrial output negatively affects unemployment as expected.6 In panel E, GDP per square kilometer is positively relatedto SO2 intensity.7

6.2. The moderating role of institutional development

The testing of the first two sets of hypotheses (H1a–c and H2a–b) is based on the results reported for Model 2 throughoutpanels A–E. We include the same sets of control variables as we did for Model 1. In addition, we include two new variables:IST and the interaction term between IST and FPtotal, for each IV regression run. The results in panels A–D show that theinstitutional development scored in the NERI index significantly strengthens the positive impacts of FDI on economic growth(b = .045, p < .10), labor productivity (b = .152, p < .01), and innovation gains (b = .487, p < .01). A closer look at the effect sizeof the results indicates that the magnitude of the positive influences of FDI heavily relies on the level of local institutions(Ellis, 2010). One unit of increase in the index leads to a reversal from a negative to a positive FDI influence on economicgrowth, a 4.5 times greater impact of FDI on labor productivity, and a 6 times greater impact on innovation gains. Thesestatistically and substantively significant results provide strong support to H1a–c, confirming that local institutions play apivotal role in stimulating and assimilating the positive influences of FDI. In locations where institutional setting is moredeveloped, FDI makes greater contributions to the local economy’s development. However, when host institutions areunderdeveloped, the benefits of FDI cannot be effectively captured.

Furthermore, we examine if institutional development also moderates the impacts of FDI on unemployment andpollution intensity (H2a–b). The coefficient of the interaction term in panel D does not reach significance, although the sign isnegative as expected (b = �0.084, p > .10). The results in panel E show strong support for the hypothesis that institutionaldevelopment significantly alleviates the environmental threats that FDI poses to the host city (b = �.420, p < .05). Takentogether, H2 is partially supported. Combined with what we found for H1, the interaction analyses points to a consistenttheme: institutions do shape, or even change the course of the impacts of FDI in a host economy. Well developed institutionswith transparent governance, efficient resource allocation and flows, consistent legal framework, and matured intermediarymarkets contribute to the better use of FDI while curtailing its potentially negative effects. To some extent, our findingsprovide an answer to the hotly debated questions of why and how the impacts of FDI vary across different contexts.

The next two sets of hypotheses (H3a–c and H4a–b) posit that the moderating effect of institutions is weaker for ethnic-linked FDI (FPel) than non-ethnic-linked FDI (FPnel). The results are presented under Model 3 throughout panels A–E of

t

D.T. Wang et al. / International Business Review 22 (2013) 450–465462

Table 3. We find that IST significantly interacts with FPnel but not with FPel in affecting economic growth (b = .052, p < .05for FPnel; b = .001, p > .10 for FPel) and innovation gains (b = .352, p < .01 for FPnel; b = .111, p > .10 for FPel), supporting H3aand H3c. Yet, both are significantly related to labor productivity (b = .109, p < .01 for FPnel; b = .103, p < .05 for FPel) with theWald test showing no significant difference (x2 = 0.01, p > .10). Therefore, H3b is not supported. Integrating these results,H3 is partially supported in that institutional development has stronger moderating effects on the economic andinnovation impacts of non-ethnic-linked FDI, but the moderation is similar for both types of FDI in terms oflabor productivity.

Consistent with H4a, results in panel D show that IST significantly reduces the impact of FDI on unemployment if it isfrom non-ethnic-linked origins (b = �.149, p < .10). However, the interaction between IST and ethnic-linked FDI is notsignificant, though in the expected direction (b = �.088, p > .10). Finally, H4b is also strongly supported in that institutionaldevelopment significantly weakens FPnel’s impact (b = �.271, p < .10) but not that of FPel on pollution intensity (b = �.074,p > .10). In all, in terms of the impacts of FDI on unemployment and pollution intensity, the moderating role of institutionsis distinct. When stronger institutions are in place, the negative impacts of FDI from ethnically distant countries areeffectively reduced. However, FDI from Hong Kong, Macau and Taiwan are less affected by the development of institutions,probably because the investors are able to bypass regulations through their local connections.

7. Discussions and conclusions

7.1. Contributions

As the first comprehensive attempt to investigate the multiple impacts of FDI in an institutional context, this paper makestheoretical, empirical, and methodological contributions to the IB literature. Theoretically, first, amassing panel data fromfour major sources covering 287 Chinese cities, we validate the double-edged effects of FDI on the local economy. Our resultswith different city-level outcomes as dependent variables provide consistent support to our argument that FDI exertssignificant positive influence on the host city’s economic growth, labor productivity, and innovation gains while posing athreat to employment and the environment. This complements and advances the existing literature which predominantlyfocuses on individual domains of the impacts of FDI. By assuming a broadened scope, this paper helps to clarify some of thecomplex influences FDI have on emerging economies.

Second, we construct an institutional framework (Henisz, 2000; Meyer, 2004; Meyer & Sinani, 2009; Peng et al., 2008), toexamine the various impacts of FDI in different recipient localities. In their encyclopedic work, Dunning and Lundan (2008)suggest that there has been some progress in understanding the determinants of MNE behavior using an institutionalapproach, yet, the approach has been less successful to illuminate the effects of MNE activities. This study fills in this gap byfocusing on how institutions, particularly formal institutions in an emerging economy, moderate the consequences of FDI.We find that by providing sound regulations and incentives, local institutions boost the host city’s ability to absorb thepositive impacts of FDI, and help curtail its potential damage. To a certain extent, this study explains why and how FDIgenerates dissimilar, sometimes contrasting, impacts across regions from an institutional perspective. As Dunning andLundan (2008) point out, FDI spillovers arise primarily from linkages formed by particular firms in particular locations. Ourstudy provides empirical evidence of the consequences of the heterogeneity in institutional structures within one countrythat influence the building of absorptive capability of local firms and upgrading of local economy.8

Third, of particular interest is our finding on the moderating effects of institutions on FDIs with different ownershiporigins. While the non-ethnic-linked FDI possesses superior resources and practices, they are less socially embedded in thehost economy. In cities with more developed institutions such as stronger support in contract enforcement and propertyright protection, non-ethnic-linked FDI exerts both stronger impacts on the positive domains and weaker impacts on thenegative domains. In contrast, ethnic-linked FDI is more similar to local firms in terms of culture, technology, andorganizational skills. Thus, they are more adaptive to the local environment. Given the institutional similarity, their impactsto local economy are likely to vary to a lesser extent. To manage FDIs discretely, policy makers should recognize thediscriminating nature, different motives and characteristics behind different kinds of FDI, so as to develop a balancedstrategy to effectively control the magnitude of the impacts of FDI across regions.

Empirically, on the basis of a nation-wide NERI index, we tease out how the variance in local institutions influences a hosteconomy’s absorptive capability in dealing with FDI impacts. Complementing the existing literature that investigated therole of institutions at the national level (Antweiler et al., 2001; Kwok & Tadesse, 1996), we capture the disparate, overlapping,and multifaceted development of institutions within an emerging economy, and confirm the effect of such institutionaldiversity on FDI impact absorption. This fine-grained within-country approach is useful given that many emergingeconomies are characterized by high contextual variations in institutional development (Hoskisson et al., 2000; Meyer,2004; Wright, Filatotchev, Hoskisson, & Peng, 2005). Our findings shed new light on how these countries can adjust theirpolicies in accordance to their level of institutional development. To make better use of FDI, policy makers need to firstimprove institutions such as infrastructures, legal systems, intermediaries, and market development, so as to stimulate andabsorb the positive FDI impacts, and at the same time, control, and dampen the negative impacts.

8 We thank one of the reviewers for suggesting this point.

D.T. Wang et al. / International Business Review 22 (2013) 450–465 463

Finally, our attempt also makes three methodological improvements over existing work. First, we use the number ofpatents approved each year to proxy for innovation activities. This provides an output-based assessment of innovationgains in a host city as compared to the input-based (R&D expenses) or derived (new products sales) measures used inprevious studies (Cheung & Lin, 2004). Second, we measure the presence of FDI as foreign output share and use thisoutput measure to predict other economic and social outputs. This helps us to avoid potential pitfalls of using the inputof FDI to predict output which may generate biased results arising from potential quality problems of Chinese statistics(Buckley et al., 2007). Third, we employ the IV methodology, using two instruments, lagged values of FDI andinstitutional development, to alleviate the potential concerns of reverse causality. Fourth, the impacts of FDI on societaldomains such as unemployment and the environment have been under-reported in IB research. Our use of city-levelpanel data allows us to investigate FDI spillovers in the societal domain. Given our finding that FDI does have negativeimplications for the host city’s employment and the environment, more research is needed to explore the nature andgovernance mechanisms of the side effects of FDI.

7.2. Limitations and future directions

Our study is limited in a number of ways. First, due to the constraint of data availability, we only have the provincialdata on China’s institutional development. If resources allow, researchers may consider conducting surveys to collectviews on local institutions in representative cities such as first-tier, second-tier, coastal, and inland cities, to obtain city-level institution measure. Although the NERI index is the most official and widely used index to measure institutions inChina, measurement reliability of the index is not available. If a national survey can be conducted to collect data on localinstitutions, researchers will have greater confidence in the quality of the institution measure and also the analysisresults. Moreover, due to the data limitation, we are unable to examine the effects of specific dimensions of theinstitutional factor such as regulatory, normative and cognitive institutions, which may play differential roles inaffecting FDI’s impacts. As the first study to employ an institution-based perspective to study FDI’s environmentalexternalities, our results suggest a promising direction. We believe future research with more refined data can morefruitfully explore this topic.

Second, the study takes an institutional perspective to examine the ambiguous impacts of FDI in five important areas.However, due to data limitation, we cannot examine other potential impacts such as corruption and income inequality.Future research can advance this study by assessing the interaction between FDI and institutional development on moredimensions.

Third, our data is at the city level; we cannot trace economic gains brought forth by FDI to particular industries within thelocal economy. Because of that, we are unable to estimate negative societal spillovers such as crowding out effects at theindustry or firm level. Indeed, future research along these lines is needed to provide insights into the mediating processes onthese micro levels.

Finally, this study focuses on how the moderating role of institutions and the different origins of FDI affect the outcomes.It will be interesting to investigate the linkages among the outcome variables. Simultaneous models could be applied tosystematically explore the structural and dynamic process of FDI impacts. For example, the mechanisms through which FDIaffects labor productivity, which in turn along with FDI affects the natural environment could be addressed by simultaneousequation models in future research. This may also allow us to study the co-evolution of FDI and institutions in emergingeconomies as pointed out by Meyer (2004) and Cantwell et al. (2010).

Acknowledgments

We gratefully acknowledge comments from Peter Buckley, Paul Ellis, Haiyang Li, Klaus Meyer, Mike Peng, Liang Shao,Dean Xu, Xinpeng Xu, Monique Wan, and anonymous reviewers. The project is fully funded by the Research GrantsCouncil of Hong Kong (GRF no. 752809) to the first author. We thank Yi Che for research assistance. All views andremaining errors are ours.

References

Acemoglu, D., Johnson, S., & Robinson, J. A. (2001). The colonial origins of comparative development: An empirical investigation. American Economic Review, 91(5),1369–1401.

Aitken, B. J., & Harrison, A. E. (1999). Do domestic firms benefit from direct foreign investment? Evidence from Venezuela. American Economic Review, 89(3), 605–618.

Alfaro, L., Chanda, A., Kalemli-Ozcan, S., & Sayek, S. (2004). FDI and economic growth: The role of local financial markets. Journal of International Economics, 64(1),89–112.

Altomonte, C., & Pennings, E. (2009). Domestic plant productivity and incremental spillovers from foreign direct investment. Journal of International BusinessStudies, 40(7), 1131–1148.

Antweiler, W., Copeland, B. R., & Taylor, M. S. (2001). Is free trade good for the environment? American Economic Review, 91(4), 807–908.Barro, R. J. (1991). Economic growth in a cross section of countries. Quarterly Journal of Economics, 106(2), 407–443.Blomstrom, M. (1989). Foreign investment and spillovers: A study of technology transfer to Mexico. London: Routledge.Blomstrom, M., & Kokko, A. (2003). The economics of foreign direct investment incentives (NBER Working paper 9489). Cambridge, MA: National Bureau for

Economic Research.Borensztein, E., Gregio, D. J., & Lee, J. W. (1998). How does foreign direct investment affect economic growth? Journal of International Economics, 45(1), 115–135.

D.T. Wang et al. / International Business Review 22 (2013) 450–465464

Bruton, G. D., & Ahlstrom, D. (2003). An institutional view of China’s venture capital industry explaining the differences between China and the West. Journal ofBusiness Venturing, 18, 233–259.

Buckley, J. P., Clegg, J., & Wang, C. (2002). The impact of inward FDI on the performance of China’s manufacturing firms. Journal of International Business Studies,33(4), 637–655.

Buckley, J. P., Clegg, J., & Wang, C. (2007). Is the relationship between inward FDI and spillover effects linear? An empirical examination of the Case of China’.Journal of International Business Studies, 38(3), 447–459.

Cantwell, J., Dunning, J. H., & Lundan, S. M. (2010). An evolutionary approach to understanding international business activity: The co-evolution of MNEs and theinstitutional environment. Journal of International Business Studies, 41(4), 567–586.

Carkovic, M. V., & Levine, R. (2002). Does foreign direct investment accelerate economic growth (Working paper). University of Minnesota, Department of Finance.Caves, R. E. (1996). Multinational enterprise and economic analysis, second edition. New York: Cambridge University Press.Chen, G., Firth, M., & Xu, L. (2009). Does the type of ownership control matter? Evidence from China’s listed companies?. Journal of Banking and Finance, 33(1),

171–181.Chen, M.-J., Su, K. H., & Tsai, W. (2007). Competitive tension: The awareness-motivation-capability perspective. Academy of Management Journal, 50(1), 101–108.Cheung, K., & Lin, P. (2004). Spillover effects of FDI on innovation in China: Evidence from the provincial data. China Economic Review, 15(1), 25–44.Christmann, P., & Taylor, G. (2001). Globalization and the environment: Determinants of firm self-regulation in China. Journal of International Business Studies,

32(3), 439–458.Cohen, W. M., & Levinthal, D. A. (1990). Absorptive capacity: A new perspective on learning and innovation? Administrative Science Quarterly, 35(1), 128–152.Copeland, B. R., & Taylor, M. S. (2004). Trade, growth, and the environment. Journal of Economic Literature, 42(1), 7–71.Dasgupta, S., Laplante, B., Wang, H., & Wheeler, D. (2002). Confronting the environmental Kuznets curve’. Journal of Economic Perspectives, 16(1), 147–168.Davies, H. (1996). High IQ and low technology: Hong Kong’s key to success. Long Range Planning, 29(5), 684–690.Davies, H., & Walters, P. (2004). Emergent patterns of strategy, environment and performance in a transition economy. Strategic Management Journal, 25(4),

347–364.Dean, J. M., Lovely, M. E., & Wang, H. (2009). Are foreign investors attracted to weak environmental regulations? Evaluating the evidence from China. Journal of

Development Economics, 90(1), 1–13.De Backer, K., & Sleuwagen, L. (2003). Does foreign direct investment crowd out domestic entrepreneurship. Review of Industrial Organization, 22, 84–97.Djankov, S., Roland, G., Qian, Y., & Zhuravskaya, E. (2006). Entrepreneurship in China and Russia compared. Journal of the European Economic Association4(2-3)

352-265.Du, J., Lu, Y., & Tao, Z. (2008). Economic institutions and FDI location choice: Evidence from US manufacturing firms in China. Journal of Comparative Economics,

36(3), 412–429.Dunning, J. H., & Lundan, S. M. (2008). Multinational enterprises and the global economy (2nd ed.). Cheltenham: Edward Elgar.Eden, L. (2009). Letter from the Editor-in-Chief: FDI spillovers and linkages. Journal of International Business Studies, 40(7), 1065–1069.Ellis, P. D. (2010). Effect sizes and the interpretation of research results in international business. Journal of International Business Studies, 41(9), 1581–1588.Ernst, C. (2005). The FDI–employment link in a globalizing world: The case of Argentina, Brazil and Mexico (Working paper). Employment Analysis Unit, Employment

Strategy Department, International Labor Organization.Fan, G., Wang, X., & Zhu, H. (2009). NERI index of marketization of China’s provinces (1997–2007). Beijing: Economics Science Press.Feinberg, S., & Majumdar, S. (2001). Technology spillovers in the Indian pharmaceutical industry. Journal of International Business Studies, 32(3), 421–437.Gao, G. Y., Murray, J. Y., Kotabe, M., & Lu, J. (2010). A ‘strategy tripod’ perspective on export behaviors: Evidence from firms based in an emerging economy. Journal

of International Business Studies, 41(3), 377–396.Gu, F. F., Hung, K., & Tse, D. K. (2008). When does Guanxi matter? Issues of capitalization and its dark sides. Journal of Marketing, 72(4), 12–28.Hale, G., & Long, C. (2006). What determines technological spillovers of foreign Direct Investment: Evidence from China (Working papers no. 934). Economic Growth

Center, Yale University.Harwit, E. (2002). High-technology incubators: Fuel for China’s new entrepreneurship? China Business Review, 29(4), 26–29.Haskel, J. H., Pereira, S. C., & Slaughter, M. J. (2007). Does inward foreign investment boost the productivity of domestic firms? Review of Economics and Statistics,

89(3), 482–496.He, J. (2006). Pollution haven hypothesis and environmental impacts of foreign direct investment: The case of industrial emission of sulfur dioxide (SO2) in

Chinese provinces. Ecological Economics, 60, 228–245.Henisz, W. (2000). The institutional environment for multinational investment. Journal of Law, Economics, and Organization, 16, 334–364.Hoskisson, R. E., Eden, L., Lau, C. M., & Wright, M. (2000). Strategy in emerging economies. Academy of Management Journal, 43(3), 249–267.Jenkins, R. (2006). Globalization, FDI and employment in Viet Nam. Transnational Corporation, 15(1), 115–142.Kathuria, V. (2000). Productivity spillovers from technology transfer to Indian manufacturing firms. Journal of International Development, 12(3), 334–369.Khanna, T., & Palepu, K. (2000). The future of business groups in emerging markets: Long-run evidence from Chile. Academy of Management Journal, 43(3),

268–285.Kokko, A. (1996). Productivity spillovers from competition between local firms and foreign affiliates. Journal of International Development, 8(4), 517–530.Kosova, R. (2004). Do foreign firms crowd out domestic firms? The evidence from the Czech Republic (Unpublished Ph.D. dissertation). University of Michigan

Business School.Kshetri, N. (2007). Institutional changes affecting entrepreneurship in China. Journal of Developmental Entrepreneurship, 12(4), 415–432.Kwok, C. C. Y., & Tadesse, S. (1996). The MNC as an agent of change for host-country institutions: FDI and corruption. Journal of International Business Studies, 37(6),

767–785.Leonard, H. J. (1988). Pollution and the struggle for a world product: Multinational corporations, environment, and the struggle for international comparative advantage.

Cambridge: Cambridge University Press.Li, H., & Zhang, J. (2007). Do high birth rates hamper economic growth? Review of Economics and Statistics, 89(1), 110–117.Li, K., Griffin, D., Yue, H., & Zhao, L. (2011). National culture and capital structure decisions: Evidence from foreign joint ventures in China. Journal of International

Business Studies, 42, 477–503.Liu, X., Parker, D., Vaidya, K., & Wei, Y. (2001). The impact of foreign direct investment on labor productivity in the Chinese electronics industry. International

Business Review, 10, 421–439.Liu, X., Wang, C., & Wei, Y. (2009). Do local manufacturing firms benefit from transactional linkages with multinational enterprises in China? Journal of

International Business Studies, 40(7), 1113–1130.Liu, Z. (2002). Foreign direct investment and technology spillover: Evidence from China. Journal of Comparative Economics, 30(3), 579–602.Low, P., & Yeats, A. (1992). Do dirty industries migrate? In P. Low (Ed.), International Trade and the Environment (pp. 89–103). Washington, DC: The World Bank.Lundvall, B. A., Johnson, B., Andersen, E. S., & Dalum, B. (2002). National systems of production, innovation and competence building. Research Policy, 31(2),

213–231.Luo, Y. (2003). Industrial dynamics and managerial networking in an emerging market: The case of China. Strategic Management Journal, 24(13), 1315–1327.Luo, Y., & Peng, M. W. (1999). Learning to compete in a transition economy: Experience, environment, and performance. Journal of International Business Studies,

30(2), 269–296.Meyer, K. E. (2004). Perspectives on multinational enterprises in emerging economies. Journal of International Business Studies, 34(4), 259–277.Meyer, K. E., & Sinani, E. (2009). When and where does foreign direct investment generate positive spillovers? A meta-analysis. Journal of International Business

Studies, 40(7), 1075–1094.Narula, R., & Dunning, J. H. (2000). Industrial development, globalization and multinational enterprises: New realities for developing countries. Oxford

Development Studies, 28(2), 141–167.Nelson, R. R. (1993). National systems of innovation. Oxford: Oxford University Press.

D.T. Wang et al. / International Business Review 22 (2013) 450–465 465

North, D. C. (1990). Institutions, institutional change, and economic performance. New York: Norton.North, D. C. (2005). Understanding the process of economic change. Princeton, NJ: Princeton University Press.Peng, M. W., Wang, D. Y. L., & Jiang, Y. (2008). An institution-based view of international business strategy: A focus on emerging economies. Journal of International

Business Studies, 39(5), 920–936.Peng, Y. (2004). Kinship networks and entrepreneurs in China’s transitional economy. American Journal of Sociology, 109(5), 1045–1074.Rama, M. (2003). Globalization and the labor market. The World Bank Research Observer, 18(2), 159–186.Qian, Y., & Weingast, B. (1997). Federalism as a commitment to preserving market incentives. Journal of Economic Perspectives, 11(4), 83–92.Rodrik, D. (1999). The new global economy and developing countries: Making openness work. Washington, DC: Overseas Development Council.Sinani, E., & Meyer, K. E. (2004). Identifying spillovers of technology transfer: The case of Estonia. Journal of Comparative Economics, 32(3), 445–466.Spencer, J. W. (2008). The impact of multinational enterprise strategy on indigenous enterprises: Horizontal spillovers and crowding out in developing countries.

Academy of Management Review, 33, 341–361.Stock, J. H., & Watson, M. W. (2003). Introduction to econometrics. Addison-Wesley.UNCTAD. (1994). World Investment Report 1994: Transnational corporations, employment and the workplace. New York/Geneva: United Nations.Wang, J., & Zhang, H. (2005). Empirical study on the effects of FDI on Chinese employment. World Economic Study, 2008(9), 13–19, (in Chinese).Wei, Y., & Liu, X. (2006). Productivity spillovers from R&D, exports and FDI in China’s manufacturing sector. Journal of International Business Studies, 37(4),

544–557.Wells, L. T. (1998). Multinational and the developing countries. Journal of International Business Studies, 29, 101–114.Wheeler, D., & Mody, A. (1992). International investment location decisions: The case of US firms. Journal of International Economics, 33(1-2), 57–76.White, H. (1980). A heteroskedasticity-consistent covariance matrix estimator and a direct test for heteroskedasticity. Econometrica, 48(4), 817–838.Williamson, O. (2000). The new institutional economics: Taking stock, looking ahead. Journal of Economic Literature, 38(3), 596–613.Wooldridge, J. M. (2002). Econometric analysis of cross section and panel data. Cambridge, MA: MIT Press.Wright, M., Filatotchev, I., Hoskisson, R. E., & Peng, M. W. (2005). Strategy research in emerging economies: Challenging the conventional wisdom. Journal of

Management Studies, 42(1), 1–33.Xu, C. (2011). The fundamental institutions of China’s reforms and development. Journal of Economic Literature, 49(4), 1076–1151.